Implements support for both individual persons and organizations as members.
Changes:
- Added MemberType enum (INDIVIDUAL, ORGANIZATION) to distinguish member types
- Added member_type column to database (defaults to INDIVIDUAL for backward compatibility)
- Added company_name field for organizations
- Made first_name nullable (required for individuals, optional for organizations)
- Implemented conditional validation:
- INDIVIDUAL members require first_name
- ORGANIZATION members require company_name
- Updated GraphQL schema with new memberType and companyName fields
- Updated all resolvers to handle new fields and validation
- Added comprehensive unit tests for validation logic
- Updated existing tests to work with new fields
- All 53 tests passing
Technical notes:
- Using memberType instead of 'kind' to avoid GraphQL introspection conflicts
- Using native_enum=False for SQLite compatibility
- Using batch_alter_table for SQLite ALTER COLUMN compatibility
- Backward compatible: existing members automatically become INDIVIDUAL type
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Add two new MCP tools for flexible GraphQL operations:
- get_graphql_schema: Expose complete schema via introspection
- execute_graphql_query: Execute arbitrary queries and mutations
These tools complement existing dedicated member tools by enabling
AI agents to:
- Discover the GraphQL schema dynamically
- Construct custom queries with specific field selection
- Handle complex queries without requiring new dedicated tools
- Work with query variables for parameterized operations
The implementation reuses the existing GraphQLClient class and
adds proper error handling for API unavailability and validation
errors. Schema introspection is formatted as human-readable text
while query results are returned as formatted JSON.
Updated README.md with:
- Documentation for both new tools
- Usage examples for schema discovery and query execution
- Guidance on when to use general vs dedicated tools
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
This change implements a Model Context Protocol (MCP) server that
enables AI assistants to manage society members through natural
language interactions.
Changes:
- Add MCP server implementation (src/mcp_server.py)
- Implement 4 core tools:
* list_members - Query all members
* get_member - Get specific member by ID
* create_member - Create new member
* update_member - Update member information
- Add mcp and httpx dependencies to pyproject.toml
- Update README with comprehensive MCP server documentation
- Create OpenSpec proposal in openspec/changes/add-mcp-server/
The MCP server connects to the GraphQL API via HTTP and provides
structured tools that abstract away GraphQL complexity.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Implement complete member management system with GraphQL API per OpenSpec proposal.
Only firstName is required; all other fields are optional with conditional validation.
Key features:
- GraphQL CRUD operations (create, read, update, delete members)
- SQLAlchemy 2.0 async with SQLite database
- Alembic database migrations
- Conditional validation (email/phone validated only when provided)
- Member fields: firstName (required), lastName, address, email, phone (all optional)
- Database seed script with sample member
- FastAPI with Strawberry GraphQL integration
Implementation details:
- Python 3.11+ with uv package manager
- Async database sessions throughout
- Proper error handling and validation
- Code quality: black, isort, ruff
- Comprehensive end-to-end testing completed
All 20 tasks from OpenSpec proposal completed successfully.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
gurixandClaude